【正文】 大學(xué)出版社，2006年1版．8曹剛，李文新．統(tǒng)計(jì)學(xué)原理[M]．上海財(cái)經(jīng)大學(xué)出版社，2006年12版．9吳曉求．證券投資學(xué)（第三版）[M]．中國人民大學(xué)出版社，2009年2版．10馬曉佳．基于材料形變理論的股票量價(jià)關(guān)系模型[D]．大連理工大學(xué)，2009（06）.11 WS Chan and YK Relation in Stocks: A Multiple Time Series Analysis on the Singapore Market [J]. Asia Pacific Journal of Management. 2000(08).12 Cheng F, Lee and Olver M Rui. Does Trading Volume Contain Information to Predict Stock Returns? Evidence from China39。論文必須按照《咸寧學(xué)院畢業(yè)設(shè)計(jì)（論文）書寫格式》的要求安排論文的版式結(jié)構(gòu)，論文的標(biāo)注必須符合學(xué)術(shù)規(guī)范。進(jìn)度及要求起止日期要求完成的內(nèi)容及質(zhì)量2010年9月 2010年10月2010年10月中旬 2010年12月中旬 2010年12月下旬 2010年12月底 2011年1月初確定課題，收集相關(guān)資料完成外文翻譯、文獻(xiàn)綜述和開題報(bào)告完成論文初稿完成論文二稿完成論文三稿、修改稿論文定稿并上交 各階段論文材料應(yīng)做到內(nèi)容充實(shí)，結(jié)構(gòu)合理；觀點(diǎn)清晰，語言準(zhǔn)確；格式規(guī)范，層次分明。審核（系、部、教研室負(fù)責(zé)人）王娜批準(zhǔn)（院系負(fù)責(zé)人）商文斌經(jīng)濟(jì)與管理學(xué)院畢 業(yè) 論 文外文翻譯譯文題目：股票的價(jià)量關(guān)系：對(duì)新加坡股市的多元時(shí)間序列分析學(xué)生姓名： 羅劍華 專 業(yè)： 經(jīng)濟(jì)學(xué) 指導(dǎo)教師： 雷紅霞 2010 年 1O月 15日股票的價(jià)量關(guān)系：對(duì)新加坡股市的多元時(shí)間序列分析W S and Y K Tse我們用Tiao和Box（1981）的多元時(shí)間序列方法來檢驗(yàn)股票的價(jià)量關(guān)系。它消除了回歸分析和傳遞函數(shù)模型的單向動(dòng)態(tài)自動(dòng)擔(dān)負(fù)的偏差。然而，明確導(dǎo)向和滯后關(guān)系的結(jié)果是混合的。雖然如此，股價(jià)與成交量不明顯的關(guān)系在確定合并成交量資料去預(yù)測未來收益是有用的。一、介紹股票市場的價(jià)量關(guān)系是學(xué)術(shù)研究者和金融市場參與者都十分感興趣的一個(gè)課題。大量的著作和經(jīng)驗(yàn)發(fā)現(xiàn)已經(jīng)產(chǎn)生，就像Karpoff（1987）的調(diào)查所概括的。在多元時(shí)間序列架構(gòu)模型化的價(jià)格和成交量的聯(lián)合是找準(zhǔn)兩個(gè)變量間有活力的相互影響的高級(jí)統(tǒng)計(jì)方法。關(guān)于研究價(jià)格與成交量關(guān)系的重要性，可能有如下原因。對(duì)技術(shù)分析來說這，此數(shù)據(jù)提供了關(guān)于未來市場運(yùn)動(dòng)的可用信息。在很多分析中都是容易接受的指導(dǎo)方針。因?yàn)樗峁┝藵撛趦r(jià)格走勢的提高警示。分析支出心理分析方法用成交量作為市場的市場看法的一個(gè)信號(hào)。第二，價(jià)量的研究可能提供股票市場內(nèi)在的微觀結(jié)構(gòu)。Clark假設(shè)資產(chǎn)回報(bào)服從一個(gè)附屬隨機(jī)過程，直接過程是成交量的累積和成交量在非過度滯后周期是獨(dú)立分布的。轉(zhuǎn)化價(jià)格改變是混合成交量作為一個(gè)混合變量的分布。這兩個(gè)變量被一個(gè)代表大量到達(dá)市場的信息的混合變量所推動(dòng)。經(jīng)驗(yàn)證據(jù)被要求去描述關(guān)于假設(shè)不同和競爭假設(shè)在股票市場的微觀結(jié)構(gòu)。他們用交易量作為流動(dòng)性指標(biāo)和爭論打成交量的股票應(yīng)該在交易時(shí)存在較少的噪音。擴(kuò)散出來，噪音可能作為價(jià)格反向的原因。在一個(gè)衡量信息包含股票交易信息的研究中，Hasbrouck發(fā)現(xiàn)安全價(jià)格交易的充分影響沒有感到突如其來，但是伴有長時(shí)間滯后。影響估計(jì)是定量的，用一個(gè)向量自回歸系統(tǒng)。真實(shí)的機(jī)制聯(lián)系交易和 的反映。向量時(shí)間序列模型被用去解決這種困難，用一種明確反映系統(tǒng)結(jié)構(gòu)和允許計(jì)算同步和滯后的影響。像這樣的例子能夠在：Richardsom，Sefcik與Thompson（1986）和Lakonishok與Vermaelen（1986）中找到。對(duì)價(jià)量關(guān)系的實(shí)證研究通常用相關(guān)分析或回歸模型。序列相關(guān)已經(jīng)被觀察到在很多回報(bào)序列中。盡管缺乏自相關(guān)結(jié)構(gòu)的相關(guān)證據(jù)，但一個(gè)實(shí)證研究表明成交量自回歸命令高于回報(bào)的（參考例子，Tes，1991）。而回歸模型則能夠把時(shí)間序列包括進(jìn)來。同樣，同期成交量的結(jié)論作為回報(bào)方程的解釋變量并不能提供一個(gè)可用的預(yù)測模型，除非建立一個(gè)完全同步方程系統(tǒng)。特別的，我們用Tiao和Box（1981）建議的模型程序。初步說明，估計(jì)和診斷檢測平衡這些大學(xué)領(lǐng)域知名的BoxJenkins方法的順序和重復(fù)步驟。一個(gè)檢驗(yàn)成交量對(duì)回報(bào)影響的替代方法是通過條件的變化。正如Lamoureux和Lastrapes（1990）的最近著作中所展示的。所以，這種方法不能消除同步偏差。本文的結(jié)果表明根據(jù)多元時(shí)間序列模型的預(yù)測勝過無經(jīng)驗(yàn)的預(yù)測和普通的時(shí)間序列預(yù)測。本文剩下部分的計(jì)劃如下。第三部分描述本文在研究中所用的數(shù)據(jù)。我們描述序列模型結(jié)果的細(xì)節(jié)，作為應(yīng)用TiaoBox理論的說明。一個(gè)參考資料在第五部分給出。 adage that volume often leads the trend of price is not supported. Nonetheless, the implicit relationship between price and volume confirms the usefulness of incorporating volume data to forecast future return. Our analysis shows that the multiple time series models outperform the univariate models in postsample forecasts.1. INTRODUCTIONPricevolume relation in equity markets is a topic of immense interest among academic researchers as well as practitioners in financial markets. 1 Since the early empirical examination of the pricevolume relation in the New York market conducted by Granger and Morgenstern (t963), research in this area has developed substantially. A large amount of theoretical works and empirical findings has been generated, as summarised in the survey by Karpoff (1987). In this paper, we attempt to investigate the pricevolume relation using a multiple time series approach. Modelling price and volume jointly in a multiple time series framework is a statistically superior approach in capturing the dynamic interactions between the two variables. As a result, improvements in the accuracy of forecasts may be achieved.The study of pricevolume relation may be important for the following reasons. First, price and volume data for stocks are generally publicly available and easily accessible. For echnical analysts these data provide valuable information about future market movements. Wall Street adages such as, It takes volume to make price move and, Volume is heavy in bull markets and light in bear markets are wellaccepted guidelines for many analysts. The study of volume is helpful since volume is deemed to lead the trend of prices, thereby offering an advance warning of a potential price trend reversal. On the theoretical level, Brown and Jennings (1989) argued that technical analysis has value in a model in which prices are not fully revealing and traders have rational conjectures about the relation between prices and signals. Analysts in favour of the psychological approach would use volume as a signal for market sentiment? A good understanding of the statistical structure of price and volume may contribute to better market timing.Second, pricevolume studies may provide insight into the microstructure of equity markets. Following the works of Osborne (1959) and Clark (1973) many theoretical models have been postulated to explain the functioning of speculative markets. Clark assumed that asset return follows a subordinated stochastic process in which the directing process is the cumulative volume, and volumes in nonoverlapping periods are independently distributed. Epps (1975) and Epps and Epps (1976) suggested models in which the variance of the price change is conditional upon the volume. Transaction price changes are then mixtures of distributions with volume as the mixing variable. Tauchen and Pitts (1983) derived a mixing variable model in which price changes and volume are simultaneously determined. These two variables are driven by a mixing variable which represents the amount of information reaching the market. As the mixing variable is serially independent, both price changes and volume are serially independent. Thus, theoretical models about pricevolume relation typically vary according to their assumptions regarding the process of dissemination of information, the rate of flow of information, the size of the market and the existence of short sale constraints. Empirical evidence is required to discriminate between differing and peting hypotheses about the microstructure of equity markets.Recently, Amihud and Mendelson (1989a, 1991) studied the effects of trading mechanism on the bahaviour of stock returns. They used trading volume as a proxy for liquidity and argued that stocks with greater volume should exhibit